We have recently identified biomarkers in DCIS biopsies that predict formation of basal-like tumors, with high accuracy, years before it actually occurs. The biomarker candidates were validated on biopsy specimens obtained from a population-based cohort of women diagnosed with DCIS, treated by lumpectomy alone and followed for twenty years [2]. These markers identify a highly aggressive pre-malignant subtype that we call "basal-like" ductal carcinoma in situ (B-L DCIS) in this application since it shares markers specific to basal-like invasive tumors. In this application, we hypothesize that by identifying additional functional, cellular, and molecular characteristics of this very aggressive basal-like pre-malignant subtype (B-L DCIS) we will be able to (a) refine their classification for risk of progression to invasive tumors and (b) develop clinically useful probes for non-invasive imaging to track these high-risk lesions in vivo. In preparation for these studies we have discovered a number of processes and potential surface epitopes specific to basal-like invasive tumors and shared with B-L DCIS and PRIMED cells. Importantly, we believe that the development of non-invasive imaging of B-L DCIS and the ability to interrogate pre-malignant cells in vivo may give us insights into the biology and risk factors critical to tumor initiation and progression, a new approach to non-invasively detect aggressive pre-malignancy in human breast cancer and the ability to monitor its'evolution as it progresses to invasive disease. We have assembled an integrated team of basic and clinical scientists to (Specific Aim 1) identify candidate prognostic markers and functions, and (Specific Aims 2-4) develop non-invasive imaging of basal-like pre-malignant lesions with high risk for future invasive tumor formation. We will: Specific Aim 1: Identify distinctive surface epitopes for human mammary epithelial and stromal cells found in B-L DCIS and characterize functional alterations that predict progression to malignancy. Specific Aim 2: Validate the efficacy and specificity of cell surface targets in vitro and in vivo using optical imaging. Specific Aim 3: Develop clinical imaging agents for detection of B-L DCIS in breast tissue. Specific Aim 4: Develop and apply functional imaging to identify basal-like pre-malignancies (B-L DCIS) that are associated with future formation of invasive tumors. These studies will generate the identification of cellular processes and molecular biomarkers that predict future invasive tumors events years before they occur. Collaboration between stromal signals and epithelial responses at the earliest stages of pre-malignancy will be elucidated. Application of these markers will allow for risk stratification of women diagnosed with DCIS and potential targets for preventive agents. Development of selected markers as tools for non-invasive imaging will open an unprecedented avenue to study early aggressive pre-malignancies and risk factors that contribute to their progression.